Developing a Domain-based Ontology using Permanent Banking Instructions

点击翻译按钮获取中文摘要

In recent years, big data of several domains has risen significantly to prominence in the world. Moreover, withthe increasing popularity of big data and machine learning algorithms, we are in the era of domain-basedinformation explosion. Therefore, it is important and necessary to develop adequate NLP and machine learningtools for the domain-specific big data. It is also important to model, classify this data and reuse the developedresources, for better information analytics and visualization. By building computational tools and linguisticresources such as a domain-specific ontology, the National Bank of Canada will be able to involve thisresource in the existing Big Data Analytics pipeline, combine and reuse it in the design of a conversationalagent and information retrieval, and thus provide more complete and intelligent product for their community.The current research project aims at creating a domain-based ontology using permanent banking instructions.Starting from domain-specific data, this collaborative project targets the pre-processing of a largedomain-based corpus using adequate tools and techniques that will consider the financial domain and thelanguage (French, in this case), the extraction of domain-specific concepts and semantics relationshipsbetween these concepts from the domain-specific corpus.The domain-specific ontology will be a valuable resource in the AI and big data analytics pipeline of theNational Bank of Canada.